A new multi-position calibration method for MEMS inertial navigation systems
Citations
524 citations
Cites background from "A new multi-position calibration me..."
...The fixed terms, and, to a large extent, the temperature varying terms, can be estimated and compensated by calibration of the sensors (see [65]–[70] for several calibration approaches)....
[...]
314 citations
226 citations
186 citations
Cites methods from "A new multi-position calibration me..."
...Finally, we define a body frame (BF), which is an orthogonal frame that represents, for example, the coordinate frame of the IMU’s chassis....
[...]
169 citations
Cites methods from "A new multi-position calibration me..."
...These methods are similar to each other [19], [20]....
[...]
References
2,536 citations
"A new multi-position calibration me..." refers background or methods in this paper
...The six-position static and rate tests are among the most commonly used (Titterton and Weston 1997) calibration methods....
[...]
...Navigation, by definition, provides the best possible estimate of a moving object in terms of its position, velocity and attitude (Titterton and Weston 1997)....
[...]
1,650 citations
470 citations
"A new multi-position calibration me..." refers background or methods in this paper
...In Shin and El-Sheimy (2002) this problem was not addressed, but it will be shown in the following sections that for MEMS sensors these errors can contribute largely to the overall position error during prediction periods....
[...]
...(a) Misaligned orthogonal sensor triad with respect to the local level frame (after Shin and El-Sheimy (2002))....
[...]
...Another issue with the multi-position calibration method for MEMS IMUs is that it is difficult to converge to reasonable bias and scale factor values for MEMS sensors without an initial rough estimate of these error values. This is due to the large parameter variation of these sensors. To provide approximate starting values for the biases and scale factors of the accelerometers, the positions closest to face up and face down configurations were used in the modified multi-position method. The reason these are only approximate values is due to the rough installation of these positions with respect to the vertical gravity vector; a perfect cube was not used as in the six-position static calibration. For the starting value of the gyroscopes, any static output can be regarded as the gyro bias since the earth rotation is negligible compared to the original gyro biases. In summary, the modified multi-position method has the following two differences than the previously published paper by Shin and El-Sheimy (2002) for use with low cost MEMS inertial sensors:...
[...]
...Angular rate tests are used for calibrating the biases, scale factors and non-orthogonalities of the gyroscopes for lower grade navigation systems. If the rate tests and the improved six-position static method are used together, one can determine all the error components even for low cost sensors. Rate tests are typically done using a precise rate turntable. By rotating the unit through given turning rates and comparing the outputs of the IMU to these references, the biases, scale factors and non-orthogonalities can be estimated. This is typically accomplished by rotating the table through a defined angular rate in both the clockwise and counter clockwise directions. A third method of calibration uses precise alignment of a multi-axis turntable to the local level frame (LLF). The IMU is mounted on the aligned turntable and then the unit is rotated through a series of accurately known angles and positioned in different orientations with respect to the LLF. This technique makes use of the gravity and earth rotation rates as references (El-Sheimy 2006). Again similar to the sixposition method, precise alignment with the local level frame is the main requirement and therefore even small orientation errors will contaminate the error estimates. Also, similar to the six-position method, LLF calibration for MEMS suffers from the fact that the earth rotation is a very weak signal and is typically buried within the sensor noise for low grade inertial sensors. This paper will explore a recent calibration method for strapdown IMU systems. This method, first described in Shin and El-Sheimy (2002), does not require precise alignment of the IMU axes and can be applied in the field for correcting changing sensor errors such as biases....
[...]
332 citations
133 citations
"A new multi-position calibration me..." refers background in this paper
...Similarly, the bias and scale factor of a superior quality gyroscope can be estimated when the average of the static value is used in equation (8) for at least 10–15 min (Hou 2004)....
[...]
...(9) Similarly, the bias and scale factor of a superior quality gyroscope can be estimated when the average of the static value is used in equation (8) for at least 10–15 min (Hou 2004)....
[...]